RODENTS OF THE GENERA RATTUS AND MUS FROM", 07THE PACIFIC ISLANDS
-By G. H. HTATE
"BUL'LETIN' OF, THE AMERICAN MUSEUM OF NATUfRAL-HISTORY
-VOLUME LXVIII, 1935 -ARTICLE III
NEW YORK February 11, 1935 AI~~~~~~~
~~~62~~
45
4I~~~~4
V711-~~ ~ ~ ~
4~~~~~~~~~~~~~~~~~L
-k~~~~~~~~~~~4 59.9, 32 R (9) Article III.-RODENTS OF THE GENERA RATTUS AND MUS FROM THE PACIFIC ISLANDS, COLLECTED BY THE WHITNEY SOUTH SEA EXPEDITION, WITH A DISCUSSION OF THE ORIGIN AND RACES OF THE PACIFIC ISLAND RAT
BY G. H. H. TATE
Although the Whitney South Sea Expedition was undertaken primarily with the object of collecting birds, a valuable series of mammals in alcohol and occasional skins and skulls have been inlcluded in the large shipments sent to The American Museum of Natural History. These animals, including bats and rodents, may probably be considered a fairly representative sample of the mammalian fauna of the principal island groups of the Pacific Ocean. I wish to express my appreciation to Mr. Gerrit S. Miller, Jr., U. S National Museum, for the loan of specimens of Rattus hawaiiensis, and for the privilege of examining the large collections in his charge, and to Dr. W. H. Osgood for a similar loan of Rattus browni and hawaiiensis and material representing other species of Rattus. Four of the five species of rodents in the collections unquestionably represent iniportations since the arrival of Europeans and Americans. Rattus exulans, however, is probably of earlier origin. This rat might be expected to conform more or less to the avian distribution patterns in- dicated below and to be present on successively fewer islands from west to east. Instead, however, it is widely and rather evenly distributed throughout most of the Pacific groups, a fact that lends support to the view advanced by Waite (1897) and others that in most, if not in all, cases it was conveyed from island to island by Polynesian man. Waite has also shown that it was frequently used for food,' has discussed its distribution at length, and has reviewed its records from the various island groups. Our own series (Table I) supplements his list of records materially. The ornithological collections of the Whitney Expedition have formed the subject of numerous papers by Dr. Robert Cushman Murphy, Dr. Ernst Mayr, and others, and important conclusions have been reached by these workers regarding the bird geography of the South Seas. In a 'See also recent opinion of Seurat, 1934, Mem. Soc. Biogeogr., Paris, IV, pp. 44-45. 145 146 Bulletin American Museum of Natural History [Vol. LXVIII recent paper Mayr (1933, p. 315) has divided the island groups (exclusive of Hawaii, New Guinea, and the Solomon Islands) into four main groups: 1.-Micronesia, extending from the Palau Islands to the Gilberts. 2.-South Melanesia, from the Santa Cruz group to New Caledonia. 3.-Central Polynesia, with Fiji, Samoa, Tonga, and northward to the Ellice and Phoenix groups. 4.-East Polynesia, including the Cook, Austral, Society, Marquesas groups, and the great Tuamotu Archipelago with its outliers, Pitcairn, Henderson, and Ducie Islands. Broadly speaking, Mayr finds that the avian populations become progressiVely more impoverished as one moves from west to east, not only between successive island groups but also in a marked degree among individual islands forming any one group, those islands most remote from the assumed source of population possessing the fewest birds. Those conclusions, though generally applicable among the Chiroptera, do not hold in the case of the cursorial island mammals, owing to the fact that they are all introduced.
Rattus exulans (Peale) Mus exulans PEALE, 1848, U. S. Exploring Exped., VIII, p. 47. Mus vitiensis PEALE, 1848, U. S. Exploring Exped., VIII, p. 49. Mus maorium HUTTON, (1878) 1879, Trans. New Zealand Inst., XI, p. 344. Mus jessook JENTINK, 1879, Notes Leyden Museum, II, p. 15. Mus huegeli THOMAS, 1880, Proc. Zool. Soc. London, p. 11. TYPE.-U.S.N.M. No. 3730, from Tahiti, Society Islands. (Selected by Stone, 1917, p. 258, from several animals from a number of islands listed by Peale under his description of exulans.) TYPE LoCALITY.-Tahiti, Society Islands. The above synonymy shows that no less than five names have been provided for Pacific Island rats as they have been successively discovered on various islands. Usually the localities have been widely separated from one another: exulans, Tahiti; vitiensis, Fiji; maorium, New Zea- land; jessook, New Hebrides; huegeli, Fiji. But note that vitiensis and huegeli are both from Fiji and con- sequently the latter is likely to be a pure synonym of the former. Hawaii- ensis Stone from Popoia, Maua, Hawaiian Islands, and micronesiensis Tokuda from Ponape, Caroline Islands, are probably both distinct from exulans-type rats. MATERIAL EXAMINED. (1) The considerable series' of 97 speci- mens (95 in alcohol, but the skulls extracted and cleaned) obtained by
'For full list of the Whitney collection see Table I. 1935] Tate, Rattus and Mus from the Pacific Islands 147 the Whitney South Sea Expedition from the principal island groups of the Pacific Ocean, the major part of which, however, is from the Marque- sas Islands and the Tuamotu Archipelago. (2) Material representing Rattus hawaiiensis and Rattus browni kindly loaned by the authorities of the U. S. National Museum, Washington, and the Field Museum of Natural History, Chicago. (3) The large comparative series of Rattus in the collections at the U. S. National Museum and the U. S. Biological Survey, examined at Washington. ANTHROPOLOGICAL RELATIONSHIPs.-I am informed by Dr. H. L. Shapiro of the Department of Anthropology of this Museum that within historical times it has not been customary in Polynesia to carry rats about in canoes, to confine them in cages, or to use them as food, though in Melanesia rats are hunted and eaten with some frequency. I inquired concerning this point because, owing to the very wide distribution of exulans, it seemed to me likely that the old time Polynesian navigators deliberately conveyed specimens of the rat from island to island with them. But Dr. Shapiro is disinclined to admit this, preferring the idea that the rats secreted themselves among the stores of coconuts and other foodstuffs on board the very large canoes used for long journeys, and were thus in a sense distributed adventitiously among the islands. Furthermore, Dr. Shapiro considers that although there are recorded many instances of rats being eaten by men, if the animals were at times kept in cages and fed, as would be necessary if their transportation were in any way customary, these facts would be known to ethnological students. Also the principles of random dispersal, even though man were the unconscious agent, ought to have operated, and, in consequence, had the distribution of these rats been of such accidental type, we ought to be able to observe from west to east a gradual decrease both in the proportion of island groups reached by them and in the percentage of islands colonized in any given group. This phenomenon though definitely observable in other organisms, for example among birds (Mayr, 1933), is not noted in the case of the exulares rats. Dr. Shapiro, while observing that a number of theories exist as to the human colonization tracks over the Pacific, states that a broadly accepted view allows two main lines of drift: One entered by way of the Caroline, Marshall, and Gilbert Islands to Samoa, Tonga, and Fiji, whence they were distributed throughout Polynesia. The other line came into Polynesia via New Guinea and the Solomons, to Fiji, Samoa, and Tonga, which served as a center for later migrations. The Maoris represent one of the early waves and are known to have migrated to New 148 Bulletin American Museum of Natural History [Vol. LXVIII Zealand from the Cook Islands. Besides the above, there is believed to have been communication between Hawaii and the Society Islands, and possibly the Marquesas. Since they must have been carried by early man, the relationships of the rats ought to follow a dispersal pattern substantially similar to that of man. But as shown below this is not entirely so, for though man crossed the gap from the Solomon Islands to the New Hebrides, the rats appear not to have done so. TAXONOMIc DIscussIoN.-In order to work out the relationships of the exulans rats, it is necessary to secure topotypes of the named forms. Three of these we now have, namely, A. M. No. 48204 from Tahiti (exulans) and A. M. No. 69264 from Fiji (vitiensis and huegeli), unless there be two forms on Fiji. The type locality of jessook is Tana, New Hebrides, from which island we have no collections, although material either from Malekula or from Efate may represent it. No example of maorium of New Zealand exists, so far as I can learn, in this country. Rattus hawaiiensis Stone (1917) and Rattus micronesiensis Tokuda (1933) are considered to be distinct species. Before discussing the Whit- ney series in detail, I wish to note certain characters of the topotype of exulans.
DESCRIPTION OF TOPOTYPE OF exulans (A. M. No. 48204, FEMALE) FROM TAHITI 1.-Palatal foramina rather narrow, slightly rounded behind. 2.-Zygomatic plate moderately rounded anteriorly. 3.-Molars moderately broadened, longer than hawaiiensis, less broad than browni. The tubercle of m2 (commonly found in browni) absent (incipient in the majority of specimens from other islands). 4.-No reduction of the posterior loph of m3 (as in raveni and eurous). 5.-Median pterygoid fossa slightly lyrate, about equaled in width by browni, hawaiiensis, and the larger Philippine species, wider than in raveni. 6.-Alisphenoid bar to squamosal moderately thickened as in raveni, eurous, basilanus, vigoratus, and luteiventris (thinner in vulcani, querceti, and browni), (very thick in hawaiiensis). Not much removed from bulla. The type skin of exulans in the U. S. National Museum has the hind foot now measuring 25 to 26 mm. Because the exulans rats have been preserved in alcohol, their varia- tion has been studied mainly through the skull and dentition. A set of measurements was first drawn up (Table VII) from which it was hoped Length of skull in mm. 28 29 30 31 32 33 34 35 36 37 38 0~E "4)u Q 3 O 2 o 1 hawaliensis , 4.
vi c chiefly young 7 0. 6 x 5 x -_ 0- 4 3 exulans \ X x 0 2 x X OO x x x x x z X x )X.OO_2 )O)OWQooQoCX) x x x x x
Hawai,i an Caroline Marshall N. Caledonia E Tuamotu II W.Tuamotu Austral Juvenile N. Hebrides Tonga Juvenile 5amoa Ellice Phoenix Cbristmnas y. Marquesas Society
Fig. 1. The upper and lower curves represent hawaiiensis and exulans re- spectively. In the lower part of the figure the curves are broken down into their geo- graphical components. Single dots represent single specimens; vertical dashes indicate two or more individuals having identical skull lengths.
149 150 Bulletin American Museum of Natural History [Vol. LXVIII to find some clue to the phylogeny of the exulans-type rats in relation to their geographic distribution. For this table the sexes were at first grQuped separately, but differences due to sexual difference were found to be largely negligible, age differences being far more important. Measurements published by former writers were incorporated into the table, includingf those given for hawaiiensis by Miller and for micronesien- sis by Tokuda. The best that can be concluded from a study of general measurements is that the eastern islands of the Tuamotu Archipelago and a few other islands seem to be populated by a form definitely larger than the rats of the remaining island groups.
PROPORTIONS OF THE AUDITORY BULLAE.-The characters of the bulla were next considered, and these, I believe, furnish us data of taxonomic importance. Only fifty specimens were found to be fully adult and intact, and thus suited for measurements of the auditory bullae. Four measurements were taken, from which the length and breadth of the bulla and the length of the bony meatal tubercle were computed (see diagram, Fig. 3): (1) lengths of bulla, (2) width across the two meatal tubercles, (3) width across anterior swollen parts of the two bullae combined, and (4) width between inner inflated walls of two bullae. By deducting (3) from (2) and dividing by 2 the width of one bulla was obtained, and by subtracting (2) from (1) and dividing by 2 the exten- sion of one meatal tubercle lateral to the bulla was arrived at. Percent- ages of width to length of bulla and of meatal extent to width of bulla were calculated, but these, by concealing the absolute dimensions, ob- scured rather than clarified the situation. STUDIES OF BULLAR OF exulans-TYPE RATS IN THE AMERICAN MUSEUM OF NATURAL HISTORY
Width Percentages Width Width apart Width Length Meatal Width Meatus A.M.N.H. meatus bullae bullae bulla 2-3 bulla ext. 1-2 Length Width No. (1) (2) (3) (4) (5) (6) bulla bulla
NEW HEBRIDES: Malekula 99755 13.2 9.1 3.5 5.62.8 5.7 422.05 49.1 73 99760 13.1 8.9 3.2 -j2.85 5.9 t22.1 48.3 77 Efate 73623 13.4 9.2 3 .0 6.23.31 6. 1.Y22.1 50.8 64 TONGA: Kelefesia 99744 13.7 9.8 3.5 6.333.15 6.1 'h 95 51.6 62 Teleketonga 77946 14.8 10.5 3.9 6'63 3 6.5 -32.15 50.9 65 t -77947 14.9 10.6 3.9 6273.35 6.5 .322.15 51.6 65 PHOENIX: Hull 68740 13.5 9.5 3 .0 6.53.25 5.9 - 2.0 55.0 62 SAMOA: Rose 68742 14.2 9.4 3.2 6-23.1 6.0 4'82.4 51.7 78 68747 13.9 9.2 3. 1 6.13.05 5.9 4-72.35 51.7 .77 FIJI 69264 13.4 9.7 2 .9 6.83.4 6.0 3-71.85 56.6 54.4 SOCIETY: Moorea 60643 14.0 9.9 3.3 fiW63 3 6.3 '212.05 52.4 62 Scilly 60785 14.2 10.5 3.2 7-33.65 6.5 -71.85 56.2 50 Tahiti 48204 13.2 9.2 3.0 6-23.1 5.9 -02.0 50.8 64.5 AIUSTRAL: Tubuai 48206 13.5 9.4 3.2 623.1 5.8 V2.05 53.4 66 Rapa 99796 14.4 9.9 3.4 Y3.5 6.1 4 53.3 69 MARQUESAS: Nukuhiva 63774 14.1 9.8 3.0 6...83.4 6.7 -32.15 50.7 63 Hatutu 99799 13.4 9.4 3.2 6.23.1 6.1 4Y02.0 50.7 64 99717 13.6 9.8 3.4 6.43.2 6.0 3.81.9 53.3 60 Fatuhiva 99720 13.4 9.8 3.6 623.1 5.9 --61.8 52.5 58 ?Island 63776 13.6 9.5 3.4 6"3.05 6.0 4'2-2.05 50.8 67 Fatuhuku 99800 13.4 9.4 3.0 6.43.32 5 .9 4'02.0 54.3 63 TUAMOTU: Tureia 99792 13.6 9.7 3.3 6 3 .2 6.2 '12Y1.95 51.6 61 Ahunui 99793 13.0 9.0 2.7 62533.15 6.1 4'02.0 51.6 67 99794 13.3 8.8 2.9 %92.95 6.2 4Y52.25 n 47.6 76 Aratika 99718 14.6 10.1 4.0 6-13.05 6.0 452.25 |i 50.8 74 ic 99798 13.1 8.9 3.0 3 92.95 6.1 4222.1 47.6 71 Raraka 99763 13.3 9.1 3 .0 6.13.05 5 9 4.22.1 51.7 69 99764 13.6 9.6 3.0 6.63.63 6.1 4.02. 0 54.1 61 Henderson 99789 14.3 10.0 3.3 6j73.35 6 .6 32 .15 50.8 64 99790 14.3 10.1 3.8 653.15 6.5 422.1 48.5 67 99788 14.3 10.1 3.6 6253.25 6.3 4.22.1 51.6 65 99783 14.3 10.1 3.4 6.73335 6.5 422.1 51.6 63 99782 15.0 10.5 3.5 7.03.5 6 .7 4"2.25 52.3 65 99773 14.5 10.3 3.4 6293.45 6.5 4W22.1 53.1 61 Hiti 60645 13.6 9.3 2.9 6.43.2 6 .3 432 .15 50.8 67 Fakarava 99725 13.7 9.4 2.9 6%53.25 6.1 4.32.15 53.3 66 Apataki 99703 13.4 9.3 3.2 6e13 05 5 .8 4212 .05 52.6 67 Ducie 99784 13.8 9.6 3.1 625,3.25 6 ° 4-22.1 54.2 65 "i 99785 13.7 9.3 3.0° -33.15 6 .1 42422.2 51.7 69 CHRISTMAS ISLAND 48212 13.1 9.4 2.9 62L53.25 5 .8 371.85 56.0 57
151 152 Bulletin American Museum of Natural History [VOl. LXVII]
SUMMARY OF THE MEASUREMENTS OF THE AUDITORY BULLAB TO SHOW POSSIBLE RACIAL STRAINE
Bulla Bulla Meatus Percentage Percentage Racial strain of exulans Length Width Width Width bulla Meatus Length bulla Width bulla
Henderson 6.1-6.7 3.05-3.5 2.15-2.25 48.5-54.2 61-74 (long) (broad) (long) Ducie 6.0-6.3 3.05-3.3 2.0-2.2 50.8-54.2 61-69 (long) (broad) (long) Raraka 5.8-5.9 3.05-3.1 2.05-2.1 51.7-53.4 66-69 (short) (broad) (long) Ahunui 6.1-6.2 2.95-3.15 2.0-2.25 47.6-51.6 67-71 (long) (broad) (long) Tahiti (exulans) 5.9-6.2 3.1-3.25 1.8-2.0 50.7-55 58-64.5 -(long) (broad) (rather short) Scilly atoll 6.5 3.65 1.85 56.2 50 (very long) (very broad) (short) Fiji(vitiensis and huegeli) 6.0 3.4 1.85 56.6 54.4 (rather long) (broad) (short) Samoa 5.9-6.0 3.05-3.1 2.35-2.4 51.7 77-78 (rather long) (broad) (very long) New Hebrides 5.7-5.9 2.8-2.85 2.05-2.1 48.3-49.1 73-77 (short) (narrow) (long)
TOTAL LENGTH OF SKULL.-Due to intrusion of growth factors the findings from the graph (Fig. 1) are likely to be less significant than in the case of the molar tooth row. It serves to show, however, that there seem to be rats of two sizes on the Tuamotus, one lot on the western and central islands ranging in skull length from 33 to 34.5 mm., and another in the east varying from 35 to 37. In the Australs the normal length is around 34 mm., but a single male reached 38.1 On the Marquesas, males, probably of separate races, vary from 33 to 36 mm. On the Fiji groups both sexes reach skull length between 30 to 36.5 mm., and on the New Hebrides males vary from 34 to 35 and females from 29 to 33. The Hawaiian rats include lengths from 30 to 33, the only male measuring 32. LENGTH OF THE UPPER MOLAR TOOTH Row.-Comparison of the relative lengths of the upper molar tooth row (drawn up from the general table of measurements) demonstrated the fact that, in the main, rats 'The basal length of adult Rattus alexandrinus equals 38 mm. or more and of Rattus norvegicu8 exceeds 45 mm. Length of upper molar series in mm. 2 48 49 50 54 52 53 54 55 56 57 58 59 6° 6' 62
8 7 )6 X 5 u) 4 l; 3 02 hawaiiensis x z ,l
17 i516 14 x Is e? 12 exulans x E 6 06a7 x 4 .2 z t Hawaii (from Miller) Caroline (from Miller) Marshall (from Tokuda) N.Caledonia I. .1 I I I E.Tu amotu W.Tuamotu I I Austral __ 1N Hebrides T onga -J 5 amoo ,Ellice Fiji Phoenix Christmas ___J Marquesas Society * From Jentinck + Frorn Waite
Fig. 2. The upper and lower curves show clearly the difference in tooth dimen- sions between hawaiiensis and exulans. The sexes are disregarded. In the lower half of the figure component parts of the curves are shown, dots indicating single speci- mens, vertical dashes two or more individuals with similar tooth row lengths.
153 154 Bulletin American Museum of Natural History [Vol. LXVIII of the various island groups have the length of the tooth restricted within characteristic limits which often overlap those of neighboring island groups but are seldom identical (Fig. 2). The chart shows, however, not only shifts in the tooth row average as between island and island, but it also brings out to some extent the fact that the rats of a given island group are not homogeneous, and to that extent supports the sug- gestion obtained by studying the bullae. Thus in Tokuda's Caroline Island rats we seem to have but one form, but in his Marshall Island specimens the tooth lengths fall into two groups, one centered about 5.2 mm. and the other about 5.8. Similar groupings can be observed in our series for a number of the island groups, such as east and west Tua- motu, Austral Islands, Marquesas, and the New Hebrides. The plotting of Miller's figures for hawaiiensis shows a distinct shift of the curve to indicate a shorter tooth row, the mode for both males and females being located at 5.2 mm., and the highest measurement being 5.4.
STRUCTURAL CHARACTERS OF THE UPPER MOLARS.-Search under the binoculars for new characters of the teeth such as accessory tubercles, which might be susceptible to a taxonomic interpretation, proved negative. Among the 50 suitable specimens examined (selected for comparatively unworn dental condition, not the same 50 as were used in the study of the bulla), accessory tubercles between the normal lophs of three main tubercles were found as follows: an internal tubercle be- tween lophs 1 and 2 of ml (6 examples); an external tubercle between the same two lophs of ml (4 examples) (Fig. 3, A.M. No. 99767-a); an external tubercle between lophs 2 and 3 (2 examples); and finally, an internal tubercle between lophs 2 and 3 (1 example). The occurrence of the above tubercles appeared to be purely adventitious and could not be correlated in any way with distribution or genetic strain. In m2 an external tubercle is usually developed weakly in front of the external cusp of the main (2d ?) loph (Fig. 3, tubercles marked b). Rather more rarely (in 11 examples) it is strongly developed, as in browni of New Guinea and the Solomon Islands, and in 8 specimens it is very weakly developed or absent. In m3 the posterior tubercle, though drawn out from side to side, is commonly entire, but in 11 skulls it showed partial fission and in one (A. M. No. 68742) it had divided in two (Fig. 3, d.) None of these characters was of assistance from the standpoint of the taxonomy of the groups of exulans. TABLE TO SHOW ACCESSORY TUBERCLES OR DENTAL PECULIARITIES OF Rattus exulans
j ~~ml M2 M3
External Tubercle IExternal IrInternal Anterior to 1st loph 3d loph
la
¢ - _ _ _ _ - 1- ¢ _ 1 - - _ _ 1- - - > -~0
Lalona~774~ 0 + +
itora 060643 + + SAMOA I +~~| +I) | | it 99757 + + Torres, Hiu 79992 + broad- ened Efate 73623 + + TONGAMaleula 99748 Lalona 77947 + +++ Teleketonga 77946 ++ + KelefesiaSit+ 997859+77945 + ++++ it 73624 + + PHOENIX NEWHull EBRIESS.68740 + + SAMOA ManuaTiaara 99760868748 + + + + Rose 68747 + + it 68742 ++ SOCIETY Scillv 60785 + + + Moorea 60643 + + Tahiti 48204 + + AUSTRAL Raivaivae 48210 + + Rapa 48209 + ± it 99787 + + Is 99796 + + MARQUESAS Tiakara 99768 ±+ +
155 TABLE TO SHOW ACCESSORY TUBERCLES OR DENTAL PECULIARITIES OF Rattus exulans-(Continued)
ml M2 M3
External Tubercle External Internal Anterior to 1st loph 3d loph
c C" ceI m 00~~~~~~~~~~~~~~"
+ ~ ~~~~~~~ "0 MARQUESAS Tiakara 99769++ Nukahiva 63774 + Traita 99767 + " + Fatuhiva 99720 + " TIJAMOTU it +~~ Hita 99761 + + + it 99762++ Tepoto 99704++ " 99716+ + + Ariteka 99798+. + - 49 99718+ " 99765 + Ahunui 99794 + + Tiakara 99769 + + Tureia 99793 + + 99716 + + TaengaAedritek 997981+99766 MariaDucie 99718499777 + + + it 99765 Rt 997761 + + + + Hendersonc 99762 + + TepotoTureia 99793199704 + + Raraka 99776 + +
99770 + CHRISTMAS 9 + ISLAND 482123 + +
156 19351 Tate, Rrltus and Mus from the Pacific Islands 157 It should be mentioned that the teeth of these rats have been affected by caries in several cases. I believe that this condition has only been previously reported for Rattus rattus, R. norvegicus, and Mus musculus.
d d
A.M 48204 U.S.N.M. 2 36899
IMd d
A.M. 9 9 7 59 F. M.31 83 4
A.M. 99767
A.M. 692 64 A.M.48204
Fig. 3. Right upper tooth rovvs of A. M. No. 48204, topotype of R. exulans; A. M. No. 99759, exulans from Malekula, New Hebrides, to show partly divided posterior loph of M3 (d); A. M. No. 99767, exulans with accessory cusp (a); A. M. No. 69264, topotype of R. vitiensis and of huegeli; U. S. N. M. No. 236899, R. hawaiiensis; F. M. No. 31834, R. browni. Auditory bullae of topotype of exulans, A. M. No. 48204, to illustrate measure- ments taken for comparison of dimensions of bulla and meatus (p. 150).
The carious condition occurs in four specimens, three from the Tuamotu and one from the Marquesas group. Comparing the dentitions of unworn specimens of hawaiiensis, the exulans rats, and browni, I find the individual teeth of browni longer and narrower and those of exulans broader and shorter than is the case with .158 Bulletin American Museum of Natural History [VOl. LXVIII
hawajiensis. Peculiarities of individual teeth of exulans are shown in the previous table and line drawings (Fig. 3) but, as is indicated, these peculiarities are often far from being constant and in most cases represent merely individual variation. The inferior teeth of all three species are remarkably alike.
COMPARISON OF UPPER MOLARS OF Rattus exulans AND ALLIES browni exulans hawaiiensis
Ml A tendency to develop Rarely such tubercles as Without tubercles of accessory tubercles be- in browni. No anterior browni and exulans. No tween the external ends cingular tubercle. Rare- anterior cingular tuber- (tubercles) of each of the ly an accessory tubercle cle. three transverse lophs. between the internal A distinct anteriorcingu- ends of lophs 1 and 2. lar tubercle. (See previous table.)
M2 With small accessory With the accessory tu- Without accessory tu- tubercle anterior to out- lbercle of browni com- bercle of browni. ermost tubercle of monly incipient. second loph. M3 Third loph commonly Third loph commonly Third loph formed by formed by a single tuber- formed of two partly single tubercle, free from cle, free from second united tubercles. Rarely second loph. Inner tu- loph. Inner and outer united with second, bercle much smaller than tubercles of second loph either by inner or by outer one of second loph. subequal in size, and outer tubercle. Inner a well-defined anterior tubercle of second loph groove between them. usually smaller than outer one.
THE GROUPS OF exulans.-Having essayed several lines of approach, namely, through proportion of bullae, length of molar tooth row, length of skull, and structure of molars, we may now try to integrate the results of these studies in order to see whether they will support one another and thus allow definite conclusions to be drawn regarding possible differentia,. tion. Below I have tried to summarize these results in tabular form. It is at once obvious that the bulla alone is susceptible of even moderately satisfactory interpretation, the other characters overlapping to a high degree and being useful only in a confirmatory capacity. One gains a distinct impression that a number of separate strains that preserve a greater or less measure of integrity are present among the islands. An 193.5] Tate, Rattus and Mus from the Pacific Islands 159 example of this type, I believe, is the form on Malekula Island, New Hebrides, which possesses a small, proportionately narrow bulla and a proportionately long meatal tubercle. Another is the larger rat of Rose Island of the Samoan group, with rather large bulla, proportionately longer and broader than that of the Malekula Island rats and also with very long meatal tubercle. Were I to reduce the number of these groups further, I would first isolate as exulans the small-sized animals with large bulla and short meatus, making the large Tuamotu rat and the Raraka, Samoan, and Fiji rats variants of it. The Malekula rats of New Hebrides I would consider more distant on account of their peculiarly small bullae and long meatus. It will at once be objected that the above divisions in practically every case are based upon very meager series. But one must use what is at hand, and I feel, in spite of a certain vagueness of the characters pointed out, that there is a considerable basis of truth in this classification. Repre- sentatives of some of the groups seem to have spread rather widely. Thus the large Tuamotu rat is present on Teleketonga of the Tonga group, on Moorea of the Society Islands and on Nukuhiva of the Marque- sas, as well as Aratika, Raraka, Hiti, and Fakarava of the west or central Tuamotu Archipelago. But its headquarters seems to be Henderson and Ducie of the east Tuamotus and Rapa of the Austral Islands, where perhaps it is exempt from intermixture and competition with the smaller exulans and Raraka rats. There is as yet no record of exulans from Easter Island, farther east, but I should expect this large race, rather than any other race, to have been taken there, perhaps by the people who carved the stone images of the Island.' Of the seven strains distinguished above, the two most strikingly separable seem to be No. 2, the large Tuamotu form, and No. 7, the variety from Malekula Island, New Hebrides. Those perhaps represent relict forms which persist only on the periphery of the principal assembly of Pacific rats. In view of the fact that the Maoris are believed to have migrated from the Cook Islands to New Zealand, it seems probable that maorium will be found most closely related to one of groups 1, 2, or 3. One can scarcely doubt that the majority of these strains of rats hybridize readily when opportunity offers, so that the greater part of the groups under consideration are likely to be of mixed origin. Since this would be most pronounced wherever communication between the natives
'See comparisons in stone workings attested by Raven, 1926, Natural History, XXVI, pp. 272-282. Q
"W °
* cq ce ao * -4-)~~ ~ CO m biC?--I 6 COo to 0' 0 A_ d 4 L m e et e co
H 0 C xo~~~~~~~~)
0 3lS_be e *Q~~~~~b "10rW dbe 0 0 z w
Co2 0Q 0E-I Ca m " V e "e Ca ;- 0 0 0 Ca0 E-4 V 0¢ Q ^~ -^; FW o b°be be ~~ be 0 ~~~~~~~~ ~ 0 0 1~ 0 > - 0 - - 0 PQ I_ _ =Co ; O I I L (a a) (a) (a. (2 ' -4- -4- C4- I3- D 3+- a) :3 O3 0 C) w s4 () w wf C) C
N 0 o - o 0 o 0 IS ;4 S
>-
-4;-0
0d adco
ziCa 1935] Tate, Rattus and Mus from the Pacific Islands 161 of different islands and island groups was greatest, we may conclude that the Society-Marquesas, or exulans division, rats having a relatively short meatus, is probably a highly mixed group which appears to have spread to the western Tuamotus, Christmas Island, the Phoenix group and (?) to Efate, New Hebrides. The Raraka and Scilly atoll animals are forms perhaps bordering on exulans proper. The Fiji Island rats are represented by a single specimen (A.M. No. 68764), which, providing there is but a single racial strain on Fiji, is topotype both of vitiensis and of huegeli. The skull, which is that of a female, is relatively heavily built, has a rather massive rostrum, and is distinct from the Tongan and Samoan rats, its nearest neighbors. The type skin of vitiensis, in the U. S. National Museum, has a hind foot that now measures 30 mm. Just as in exulans, the tail has scale hairs equal in length to one and one-half to two scale lengths, the hairs above dark, those beneath white. HAWAIIAN AND CAROLINE ISLAND FORMS OF exulans; ALso Rattus hawaiiensis AND micronesiensis.-Before turniing to the question of the origin of the rats of the Pacific, some further factors must be taken into account. These are (1) the rats from Popoia, Oahu, Hawaiian Islands; (2) the common rat of the larger Hawaiian islands, of which a dozen from each-Oahu, Maui, and Hawaii-have recently been collected by Dr. Eskey and sent to Washington; (3) the rat of Ocean Island at the extreme west of the Hawaiian chain, some 1200 miles from the main group, a considerable series of which were secured by Dr. Wetmore of the U. S. Biological Survey; (4) the " exulans " of Tokuda from the Caroline and Marshall Islands; and (5) Rattus micronesiensis Tokuda from Ponape, Caroline Islands. The skulls of hawaiiensis from Popoia are immediately to be distin- guished from any hitherto discussed by their smaller size, shorter, more widely separated tooth rows, short rounded bullae, differently formed zygomatic plate, lack of accessory dental tubercles, undivided last cusp of m3, and rather spinous fur. The mainland rats collected by Dr. Eskey seem different from those of Popoia on account of their different "make." Their pelage appears less spinous and longer haired, and, taken with other specimens from the main islands, they evince a seasonal color change from more cinnamon to more grayish, as do also the Popoia rats. Although I could discover no anatomical difference from the Popoia animals, I obtained rather a different set of measurements from those of Miller for the Popoia rats (Table VII, p. 172). 162 Bulletin American Museum of Natural History [Vol. LXVIII Ocean Island rats belong to the Pacific rat type rather than to the Hawaiian rat type. It was suggested to me recently that because Ocean Island is periodically swept by hurricanes the rat fauna would be period- ically destroyed, and in consequence the present rat colony must be lately introduced. I am permitted to quote from a letter in which Dr. A. Wetmore was good enough to express his opinion. Referring to Green Island, Ocean Island, he writes: This island at its highest point lies from twenty to twenty-five feet above high tide mark. It is surrounded by sand beaches back of which there are knolls of sand covered in the main with a dense growth of a shrub sometimes called beach magnolia (Scaevola). In the center there is a grass-grown opening of fair size. A smaller island, called Sand Island, has less elevation and no shrubby vegetation, as it is evidently swept regularly by high water. I have little doubt that hurricanes on occasion sweep Green Island . . . at such times, I presume that waves inundate the land though I have no definite records to that effect. Probably the outer reef breaks the force of the water . . . .Even though the island is swept on occasion by water, I see no reason to suppose that the small rat may not have been able to persist. These creatures climb readily and the small stock might be saved in some hollow tree trunk. That the island population of rats is replaced regularly I doubt exceedingly as even now there is little traffic to this island. It is my assumption that the present stock of rats came there long ago Writing of Wake Island north of the Marshall group, Dr. Wetmore mentions the presence of "a flightless rail peculiar to this one atoll. I assume that if these rails could exist the rats must likewise have been able to maintain themselves." The rats of Ocean Island have well-inflated bullae, a wide pterygoid fossa, more projecting zygomatic plate, and less proodont incisors than hawaiiensis. They lack a postnasal spine. Considering the exulans-like character of the rats on Ocean Island, I would expect the rats recorded by Tokuda from the Caroline and Marshall Islands to belong with the Pacific Island rats, and this view is strengthened by the presence in Tokuda's diagram (p. 81) of a notch in the posterior loph of m3, in perfect agreement with the usual condition of that tooth in the exulans-type rats. His half-tones (PI. v) are scarcely clear enough to form the basis of an opinion, but the skulls illustrated appear to me to have widely flaring meatal tubercles and thus to agree with my groups, 4 (Raraka), 5 (Samoa), and 7 (New Hebrides). But the bulla appears wide. thus eliminating group 7. In all likelihood Tokuda's aniinal will prove to be a mixed group allied to the Samoan and Tuamotu rats. The wide variation in such measurements as length of tooth row (5.0-6.3 mm.) and length of bulla (5.5-7.0 mm.) indicate a mixed strain. 1935] Tate, Rattus and Mus from the Pacific Islands 163 His numbers 1, 7, and 10, all from the Marshall Islands, suggest the large Tuamotu rats. Micronesiensis from the character of its "bullae narrow and high," recalls one of the smaller members of the concolor group of the Philip- pines or raveni of the Celebes. Its wide palate is similar to that of hawaiiensis (which can otherwise be ruled out on account of the bullae). It seems from the plate to have a rather widely extended meatal tubercle. There is in the Philippine Islands a group of small species-leucophaeatus, ornatulus, pantarensis, and basilanus-to which it is possibly allied, and it tallies with this group in foot length, though its tooth rows are very slightly longer. THE concolor GROUP AND RELATIONSHIPS OF THE PACIFIC ISLAND RATs.-When a comparison of hawaiiensis, micronesiensis, and the several races of exulans-type rats is made with other members of the concolor group to which they unquestionably belong, very little hint of relationship to any one species more than to another can be gained. And indeed, an analysis of the whole group is needed before their position can be satisfactorily determined. The "concolor" group of Rattus, so far as it can be characterized at all, is an assemblage of quite small, brownish or grayish rats with whitish or buffy under parts, commonly possessing a greater or less development of spinous hairs mixed with fur and with the tail usually but not always slightly shorter than the body. Females of this group invariably have two pairs of pectoral and two pairs of inguinal mammae. No trace of opposability in the hallux can be noted. The skull and teeth appear practically unspecialized. The length of the skull ranges from about 29 to 38 mm. The concolor group converges upon several South Asiatic and Malaysian groups of rats among which may be mentioned cremori- venter, fulvescens ( =jerdoni ?), and asper. The geographical ranges of these groups appear to be respectively as follows: concolor group:-Tenasserim, Malay States, Sumatra, Java, Borneo, Philip- pines, Celebes, New Guinea, eastward from these lands over almost the whole of the Pacific Islands and New Zealand. cremoriventer group:-Siam, Jalor, Annam, Cochin, Tenasserim, Malay States, Sumatra, Java, Borneo, Bali (absent from the Philippines). fulvescens group:-Fukien, Yunnan, Indo-China, Burma, Nepal, Sikkim, Siam, Tenasserim, Johore, Sumatra, Java, Borneo, Celebes (also absent from the Philippines). 164 Bulletin American Museum of Natural History [Vol. LXVIII The following notes are drawn up from three specimens of true concolor, U.S.N.M. Nos. 152189, 240541, and 111981, the typical species of the group, which inhabits Siam, Tenasserim, and the Malay region. 1.-Skull medium in size, greatest length 29.5 mm. Zygomatic breadth 14.5 mm. 2.-Palatal foramina slitlike, rather pointed at both ends, 5 mm. (up to 5.9) or slightly less in length, occupying two-thirds of diastema (7.5 mm.), posterior end level with front of ml root. 3.-Muzzle narrow, rather short, front of zygomatic plate distant from anterior edge of premaxilla 6.6 mm., and from front of orbit (below) 9.6 mm. 4.-Zygomatic plate thrown forward, rounded; infraorbital foramen well opened above. 5.-Molar series (alv.) approximately 5.0 mm. 6.-Bullae large and rounded, 5.6 mm. X 4.0 mm. (to meatus), 2.6 mm. apart across basioccipital. 7.-Median pterygoid fossa of moderate width, not lyrate in form; sides nearly parallel; palatal edges rounded to meet in a slight postnasal spine. 8.-External pterygoid wings of alisphenoid slightly variable (?); broader in specimen from Selangor; narrower in one from Indo-China (Luang Prabang). 9.-Occiput not greatly produced. 10.-Well-marked temporal beading. 11.-Posterior edge of mandible between articular and angular processes well excavated; and a well-developed coronoid spine. Skin crisp-haired with short spines. Length of ears moderate, 13-14 mm.; feet,24,23.5,25.5 (= 23.5-25.5). Tail slightly longer than head and Tail "c122" body. H or, remeasured on the study specimen, Head and body "94" 109. Tail hairs three, subequal (median slightly longer), about two 116 scales in length, black dorsally and white ventrally. Colors dull grayish brown above, dull brownish buffy beneath, the ventral hairs very short; gray bases relatively long; brown tips occa- sionally worn away. Throat occasionally without gray bases to hairs. Feet pale buffy brown. Juveniles with pelage rather long. The group as a whole is readily separated from allied groups, cremoriventer, and asper. Distinguishing characters of groups of Rattus allied to concolor: 1.-Concolor, with large rounded bullae and narrowed basioccipital; palatal foramina slitlike to rather rounded behind; moderate to short tail; fur spinous or not so; feet moderate in length. 2.-Asper, with small bullae and wide basioccipital; short rounded palatal foramina; short tail; heavily spinous pelage; long feet. 19351 Tate, Rattus and Mus from the Pacific Islands 165
3.-Cremoriventer, with small bullae and wide basioccipital; moderately short- ened, but wide palatal foramina; extremely wide pterygoid fossa; very long, well- haired tail; moderately spinous hair; short feet. The New Guinea species browni which is currently placed in the concolor group is characterized by a well-developed extra cusp in m2, a short tail, heavy spines, and a lack of soft fur. Species adjoining the Pacific area referable to the concolor group, with the regions occupied by them, are as follows: PHILIPPINES: 1.-Small, heavily furred, nearly spineless, calcis, negrinus 2.-Large, moderately furred, querceti, vigoratus, vulcani, todayensis, mayonicus 3.-Small, short furred, leucophaeatus, ornatulus, pantarensis, basilanus 4.-Luteiventris BORNEO: concolor subspecies trachynotus ? schuitemakeri ? CELEBJES: raveni raveni, raveni eurous FLORES: uichmanni ? NEW GUINEA: browni concolor lassacquerei It will be noted that in all regions, except in the Philippines, species are few (the mainland, Sumatran and Javanese animals are chiefly forms of concolor and ephippium). This is perhaps due to the relative develop- ment of competing species of the cremorriventer and fulvescens groups. In the Philippines, on the contrary, where those groups are absent, the concolor group has speciated freely. The accompanying map (Fig. 4) indicates the probable course of divergent evolution in the concolor group. Browni of New Guinea and the Solomon Islands appears to be a peripheral species which has failed to enter North Australia. At least I have failed to find it or any representative of the concolor group among the collections of Raven and of Hoy. Of wichmanni I know nothing beyond the descriptive notice. Considerable evidence is adduced by Dickinson (1928, pp. 273-289) and others to the effect that the Philippines were rather recently (late Tertiary) linked with Palawan and Borneo, and Borneo with Java, Sumatra, and the Malaysian mainland, Celebes having been more re- motely connected. It is probable that the present Philippine radiation 166 Bulletin American Museum of Natural History [Vol. LXVIII was initiated before severance of these islands had been completed. Some evidence is available in confirmation of this view in the presence of what I believe are representatives of the large Philippine mountain section of concolor (vigoratus, todayensis, etc.) on the mountains of Java (specimens in U.S. National Museum). Not improbably the section is represented on the higher mountains of Borneo, though as Raven's Borneo collection is from the lowlands only, and I have been unable to relate any of Thomas's species to the mountain concolor group, I am uncertain on this point. In Raven's mountain collection from the
Fig. 4. Schematic view of the distribution tracks of the concolor group (except the Pacific Island forms) from a postulated place of origin in the Burma-Siam area. Celebes there is no representative of the concolor group, raveni raveni aind raveni eurous being both lowland species. I am inclined then to look upon this high mountain section of the concolor group of rats as a relict fauna, whose spread into the lowlands is opposed by the lowland concolor species leucophaeatus, etc.; luteiventris; and true concolor and ephippium with their geographical races. These mountain species may be viewed as peripheral to the lowland forms in an altitudinal sense, just as brotwni of New Guinea is peripheral in a geo- graphical sense. Finding that exulans matches the Philippine mountain forms rather closely, I am inclined to consider the several Pacific members of the 19351 Tate, Rattus and Mus from the Pacific Islands 167 concolor group, excepting micronesiensis, as belonging to this same grade of relationship. I would suggest a similar concentric arrangement in the Pacific (map, Fig. 5), the newest arrival and at present dominant portion being perhaps the typical widespread exulans, with an older remnant in the New Hebrides, another (hawaiiensis) on Popoia, and a third represented by the large Tuamotu rat, Tokuda's large Marshall Island rats and (possibly) maorium. I would thus derive exulans and hawaiiensis from the Philippine-Borneo-Java region in the form of many successive waves probably arising from stocks already differentiated from each other. I am not inclined to believe that any one of them has passed through New Guinea. Rattus rattus alexandrinus (Geoffroy) Mus alexandrinus GEOFFROY, 1803, Cat. Mammif. Mus. Nat. d'Hist. Nat., Paris, p. 192. Recently, Aharoni (1932, pp. 178-181, 229) has shown that two forms of rat have been confused under the name alexandrinus: "a gray bellied and gray-brown backed form" of Egypt, Palestine, and Syria, which she has recognized as true alexandrinus Geoffroy; and a "white bellied, dark brown backed form " from Italy, Syria, and part of Palestine, to which she has applied the name frugivorus Rafinesque, with synonym tectorum Wagner. She has considered both to be subspecies of rattus. In the present collections only true Rattus r. alexandrinus has appeared. It seems to be abundant whenever established. (See Table II for records obtained by the Whitney Expedition.) Rattus rattus rattus (Linnaeus) Mus rattus LINNAEUS, 1758, 'Syst. Nat.,' 10th Ed., I, p. 61. The black rat occurs only twice in the collection, and from localities more than 1500 miles apart. Probably this form is only moderately common on account of constant crossing with alexandrinus. In South America I have sometimes found it locally abundant in rather new clearings in the forest. (See Table III for records secured by the Whitney Expedition.) Rattus norvegicus Berkenhout Mus norvegicus BERKBNHOUT, 1769,1 'Outlines Nat. Hist. Great Britain,' I, p. 5. The Norway rat seems to be slower than alexandrinus to colonize. The Whitney collections contain only six specimens, all from the East- Polynesian2 division (see Table IV). 'From Cabrera, 1932, Trab. Mus. Nae. Cien. Nat., Ser. Zool., Num. 57, p. 264. The origin of norvegicus has commonly been attributed to Erxleben, 1777. 2See Mayr. o ¢
*S 4
0 Ca
=b
A3Q 00 B-4 to C+._
4 ) o 3 23.
- Ue
b bn 1935] Tate, Rattus and Mus from the Pacific Islands 169 Mus musculus Linnaeus Mus musculus LINNAEUS, 1758, 'Syst. Nat.,' 10th Ed., I, p. 59. The house mouse is represented in the Society, Tuamotu, and Marquesas groups each by a single specimen. It has almost certainly a much wider distribution in the islands. Revilliod' has described a new subspecies from New Caledonia and the Loyalty Islands. (Table V.) SUMMARY AND CONCLUSIONs.-Except for the ship-borne Mus musculus, Rattus norvegicus, Rattus rattus rattus, and Rattus rattus alexandrinus, the only group of rats that has colonized the Pacific Islands is the Malaysian concolor group. None of the island members of this group is identical with any living mainland, Bornean, Philippine, Cele- besian, or New Guinean species. The animals as a whole tend to be slightly larger than most of the mainland forms and reach comparatively large size in the extreme east and south of their range. Anatomically slightly specialized forms usually unworthy of nomenclatorial separation, probably representing early colonizations, are present on the Hawaiian Islands, and Tuamotu-Austral Islands, and on the New Hebrides. On the remaining groups of islands, less specialized but still divergent forms occur, the exact status of which is obscured by frequent crossing and by the probability that they originated from different racial strains arriving in the islands at different times. Since all have been carried by man, whether voluntarily or involuntarily, they must have followed his colonization tracks; yet in the case of these island rats the general laws of dispersal seem to have been interfered with, so that, contrary to the case in most organisms, no noticeable diminution of the number of forms developed or numbers of islands reached from west to east is to be per- ceived. Most of the colonization was achieved before the advent of white men; lately due to deterioration in the quality and quantity of native boat building the mixing process has fallen off. The original course from the mainland was probably from Borneo and the Philip- pines via the Caroline Islands rather than through New Guinea and the Solomon Islands.
'Revilliod, P., 1914, 'Nova Caledonia,' I, pp. 364-365. TABLE I.-LOCALITIES OF Rattus exulans WITH NUMBER OF SPECIMENS COLLECTED BY THE WHITNEY SOUTH SEA EXPEDITION
MARQUESAS ISLANDS AUSTRAL ISLANDS Apataki Rapa 5 Ariteka Raivaivae 1 Fakuhuku Tubuai 3 Fatuhiva PHOENIX ISLANDS Hatutu Hull 1 Hita Canton 1 Hivaoa CHRISTMAS ISLAND 1 Nukuhiva TONGA GROUP Roraka Kelefesia 3 Taenga Lalona 1 Tepoto Teleketonga 1 Tiakara SAMOAN ISLANDS Traita Rose 2 Uapu Tau 1 TUAMOTU ARCHIPELAGO Ofu 1 Ahunui FIJI ISLANDS 1 Ducie NEW HEBRIDES Fakarava Efate 1 Henderson Banks 1 Hiti Hiu 1 Manihi Malekula 7 Maria No data. 8 Pitcairn Tureia 97 SOCIETY ISLANDS Moorea Scilly atoll Tahiti
TABLE 11.-LOCALITIES OF Rattus rattus alexandrinus WITH NUMBER OF SPECIMEN COLLECTED BY THE WHITNEY SOUTH SEA EXPEDITION
MARQUESAS ISLANDS 1 SOCIETY ISLANDS Hivaoa 1 Talati 1 Nukuhiva 1 Tahiti 2 Toau 1 Moorea 1 Fiji ISLANDS 1 Bow Bow 1 CAROLINE ISLANDS No data 1 Kusaie 2 13
170 TABLE III.-LOCALITIES OF Rattus rattus rattus WITH NUMBER OF SPECIMENS COLLECTED BY THE WHITNEY SOUTH SEA EXPEDITION
SOCIETY ISLANDS CAROLINE ISLANDS Moorea 1 Kusaie 1 2
TABLE IV.-LOCALITIES OF Rattus norvegicus WITH NUMBER OF SPECIMENS COLLECTED BY THE WHITNEY SOUTH SEA EXPEDITION
SOCIETY ISLANDS SAMOAN ISLANDS Tahiti 1 Tutuilla 3 Moorea 1 HORNE ISLANDS Fotuna 1 6
TABLE V.-LOCALITIES OF Mus musculus WITH NUMBER OF SPECIMENS COLLECTED BY THE WHITNEY SOUTH SEA EXPEDITION
SOCIETY ISLANDS MARQUE.SAS ISLANDS Moorea 1 Hivaoa TUAMOTU ARCHIPELAGO Hiti 1 3
TABLE VI.-LOCALITIES OF Rattus browni WITH NUMBER OF SPECIMENS COLLEICTED BY THE WHITNEY SOUTH SEA EXPEDITION
SOLOMON ISLANDS New Ireland 8 skins and skulls 2 alcoholics (skulls extracted and cleaned) NEW GUINEA Sevia 3 skins and skulls
171 0 >l>)
-q- -4 4XD Ca S S PO4 S ' ff llWOcOdCaHX?N1I|tCa~COttCIC.0 114 I C-o~C X ++C 0 O~~~~~~~~~~~~~~~~~~-htO4A-Illxioo)a) t-.. a) C)t2 C _- ) bJ 0 q > U:-;D15-)o--re o00- 0 < C < < < eo< < '-''-'< < <-' < < < o U: R~~~~~~~~~~~~~~~~~f t- 00 R1 .d 00 r-N N4 e.I 4 q LoC: - co u00O-t Lo Ez~~~~~~~~~~~~~~~~L 3IUxoLO LO to tO LO LO 1 LO 1 LO LO tOtoL 0 LO LO 'N t- N Z N No -N Nt N N toN N -. Blq!putNB~~~~~~~~~~~~~~~~L LO Lio ouLo LXoLo Lao LoLo Lo Lo Lo LO LO Lo LO LO b m EX wnsXsou Jo llXpXeal~~~~~~~0 g~ 1- cs C'l eq -t cczLO c0zc r C c c>s SJOSI O O Ao x ^~~~~~~~~~~~~~~~~~ o Lo t
V M9psa 2[ DI}SUO3XZ. . . . . ql u i jv ! ix 4u . V~~~~~~~~~~~~~~~~t LO o Lo o Lo Lo L to Lo to to LotoL LOO
¢a2 ceCO~00es -- ce coLo ce coCt ee co en A bPvP4COW.!LUI0_-_
IIIsJtlur1swJ
¢~~~~40 Pu 1.
rA S pog put, pvaH|
v¢; E xa "o 'b 'b oraoDOOOO H H `b b 'lb bo b b b b b b
E- z .! amux
M~~~~~~~~~~~~Ca~~~~~~~~~~~~~~~~~~~~~~~~~~C +0 +0+ +0 +0 +0~~~~~~~~~4-D + +0- ) +.0 0 0)+00)0)0)0)+0-)0)
b.00 W i) bO 0
ocN 0 c co m 0m c OC0t= II
- 00mt.- - - - N- - C - CD N CD - - - t- --~c - -
100M00( 0 M( C (M r--l ICO 0CO m OCO0= COmC=o 0 COCOCOCOCOCOCOCOCO ~~~~~~~~~~COCOCCO~~~~~~~~ ~~~~~COCO cOCq CO C
M M10"t M 1tiqO c q D C*) cv4
. .. ~cDoc 1 t- c n - o L 10LO1- C
cck-o iq111CCO - f 0liCr- O c .~~~~~~C~ 10LOLOllr~toltD xo to lt-~ U-~ xo In in LO LO LO LO LO LO .. Lo IC) LO LO LO LO cq-4qmr-- m O-OO COCYC4 C
LOO-tm m m cli LO cq 00 co Z Nc qCD C CIDCYDmctzCtDcyzcyzCOD CfD m m M co co co m m CYD CYD m CYD m m~~~0
cq cq~~~~~~~~~~~~~~0
LO tE0O co= m~H~ + U10oA lTlTal
[ LO CO Co N CO 0 cO cN tlpooJ, ...... Molj ulqnqlputl LO LO ul to 10 to LO o co to 1 MOa Ma00s SX9IllXV?N X CC b C L
alqlpusK~~~~~~il Lo In Ul.o LO LO co LO ot14H I - i C4Hcs
. alq!pu,uN ~ . . . . . S.IOS!OUIJO0SWOOJuA19^ s>oX ~ t .
SIOSDujJoIOOI Igo 0 o0COoqll OC ro C to c.c-- o-oo-- < o- ,- - -
- O-o c 0 0C)0 0 00OcoC CD 0 CC 0 0- 0- 0_so - IOOPH co o0 * ~~~~~~~~~~~~NNl ~ ~N1 ~ Nl Hg Xpoapu PlalIH X N-C~ C
1xz 'e~~ccc oX*~ o-'o~ o a) OO DclotN CCo _l t -4 -1 11 H 4 ~q'
__~~_toN_-co-______~ t - 0 CO N ~~~~~~~~~~~~~~~~~~~OC* N ~ C EtuX8 E vE Ca ta
0)-e £0
- 0 cd0 C00 e40~~O ~ ~~~-l~~~~~~ a CC :5 Ca z X >; t = O X; > 5 ° = c~~~- 0 ~- ~ ~ ~
cq oL cz c es c es ~~~~~~~~o4......
LO LOto<<£U: ~ ~ ~ ~ ~ ~ ~ ~ ... t~ In0 0 to~ csc:DXooNXH cet co LOX eco t e e < u:~~~~~~~~~~~~~~~~~~~Cc c e£ 00 o o000 - oifOo aoL:£ CCO so ~~~OU ~(MCo ~ ~c ~ n c n )0~0ot _ _1oH " C to o 4~~~ ~ ~ ~ ~ ~ ~ ~ C CC_ oo ct X co ce.I: 0~~~00~~~~~~~~~~~~~~~ r,^4 Oq 00.C 0~~~~~~~~~~~~~2 oz ~~~~~~~~~~~~~~~~~~~z1-4 zC) Q Q QQ- U1OAA 114"I C C~ 0 bkb E: sE n S S3 S S -- S MoI u4oojL juinqlpuuW N NOO M0g~~~~~~~~~fLO0LOllXP|-r:Io LO Lr LO LO LO LO olq!pu,uW SJOSIOUI JO S94OJX lAo iun.i4so,d jo Tj4puajq ~ t's n' CD:LO ([: < q4pptojq ju4qioia4uj.I~~~~~~~~~~~~~- qXpwalaaz9B°3Z s cc cq c c c c cL I--, jn)jS jo q4Wua'I 4sa4vai3 n noC:eno Z2.0 CO ~Cn~ C: L m D LoDX looj pUI ;t C O f !HBX >k00 ccoc LO 0 r- 0 el4O o rtoL H dqC Q1 t _ *: n il t "! Cq C SP09Put' PBOH _1 C - c qCf - L-Z c t- _f _) t- -~~~~~LlNCccicCeCf o' pUe`7a0 c. o lb lb lb lb lbl o, s o, 0 b lb o aqwunX N-tt- C - em i- C01IkO C0101a0c0C0 : 4 0 CO Apas cc 0~~~~0 3 ¢L '0 * ¢zEQ S ¢ ~- s s; B CLW E-0h0 >10 _ cs c d Ca ce d fl Ml Ca 4clXXe r- 00O O0D0bo o o o o o O ss~~~~~~~S -ssS~ S S 32 00~~~~0000O~~~~~~~ 0~O0 ifC oOof~Lf1oo o o o L LO L L c c C;C; -Xl C;Cq000C 0 N0I 0 0-co0 x CD1 -. ------4 - - - - 0 0eDCS0 o L b cDO00t O ao00 ce00dq OCQ O LO LO- C O---O - -- O O C -- - OO O--4 - - - C co- ~~. LO 0 C;l 't 00 0 =C'(IN1 C)tO t- 00 CD CS CS t, 'tj 1 ~~~. . +~~~~~~~~~~~~L li< ti ll LO Lo O to co co CD = . CLo= cc u Cq00 N C J - c1Yc:_NlO - t- LO L 00 ~~. ~~~~~~~~~~~~~~~9Ocs o X eq o M " mt m " ---I e s M N c;lcse: ------0X 00 C5O--b <-----u: 00 C) ko If to In, cl 1- - i C :C)tC CS dZ d cCf e> co c) d-<' 0 m C) to m t- d 0=O C4 LO t- 0 O0 0 Ct O a)O C)LO< C 0 co m C4 m Cq - m mC_tC4 'It LO tM CY N CSN LO mO "O0+. 01 07"c lb "7 ol o0o-lb lb'iQdqC- b'b 'b 'b lb r. ob os o- o+ ---0 _- C - el .- coSCi L N t- r--q0mc 3 4 C4 m0 0DC C co O V22 0L;~~~~~~~~>k s 178 Bulletin American Museum of Natural History [Vol. LXVIII LIST OF REFERENCES AHARONI, B. 1932, Zeitschrift fuir Saugetierkunde, VII, pp. 178-181, 229. BULLER W. 1871, Trans. New Zealand Inst., III, p. 1. DICKERSON, R. E. 1928, 'Distribution of Life in the Philippines.' HUTTON, F. W. 1887, Trans. New Zealand Inst., XX, p. 43. JENTINK, F. A. 1879, Notes Leyden Mus., II, p. 15. MAYR, E. 1933, Mitt. aus dem Zool. Mus. in Berlin, XIX, pp. 306-323. MILLER, G. S., JR. 1924, Bull. Bernice Bishop Mus., XIV, pp. 1-6. OLIVER, W. R. B. 1911, Trans. New Zealand Inst., XLIII, pp. 535-539. PEALE, T. R. 1848, U. S. Exploring Exped., VIII. REVILLIOD, P. 1914, Nova Caledonia, I, Lief. 4, pp. 343-365. THOMAS, 0. 1880, Proc. Zool. Soc. London, p. 11. TOKUDA, M. 1933, Annot. Zool. Japon., XIV, pp. 79-84. WAITE, E. R. 1897, Mem. Austral. Mus., III, p. 165. WHITE, T. 1894, Trans. New Zealand Inst., XXVII, p. 245. --A4~~~~~~- 7-1~~~~~~- A-r~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A VI-~~ ~ ~ ~ ~ ~ ~ ------t - '1~~~~~~~~~~~~~~~~4 IA~~~- 74 '4 9 kZ~~~~~~~~~~+ i --t~~~~~~~~~~~~~~~4--~ ~ ~ ~ ~ ~ S5-~~~~~~~~~~~~~~~~~~~~k -"--tsr 1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~11 s-mV"~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 -~~~~~~~~~~~~~7ff7' 71-~~ ~ " <~~~~~ 4 4,~~~~~~~~~4 '3 ~~ ~ ~ . ~ ~~~~~~55¾'~~~~~~~~~~~~~~~~~"- 's ~5~ k~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~J; _41~~ ~ ~ ~ ~ ~~4 ½ % :7~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ V~4 " 4~~~~~~~~~'N1